Private cloud storage vs storage virtualisation

What is private cloud storage, and how is it different from a pool of capacity linked by storage virtualisation? Independent consultant Chris Evans explains.

Private cloud storage and storage virtualisation both allow an IT department to supply the business with data storage from a common pool of capacity. What sets private cloud storage apart is that it allows the IT department to scale customer storage, manage customer requirements and possibly charge them for what they use.

In this interview, Bureau Chief Antony Adshead speaks with Chris Evans, an independent consultant with Langton Blue, about the differences between private cloud storage and storage virtualisation.

You can read the transcript below or listen to the podcast on storage virtualisation vs private cloud storage. What are the key characteristics of private cloud storage, and how is it different to a pool of virtualised storage?

Evans: The best place to start is with a definition of those two things.

First of all, [with] private cloud storage, we’re talking about deploying hardware solutions that meet a certain number of requirements. Those would include scalability--the ability to scale that environment up or down, probably more up than down with most people.

[Then there is] the ability to have that multi-tenant, so internally you’d expect that to be serving multiple clients, and you’d want to be able to isolate them from the management and performance perspectives. And, of course, you want to charge for the usage of that environment based on usage and not necessarily of the hardware itself.

Those sorts of things abstract the deployment of the hardware into what we term a cloud deployment.

With [storage virtualisation] ... people would use a middle layer to present multiple pieces of hardware to servers in a virtualised fashion. So, for example, Hitachi [Data Systems] can virtualise other storage vendors’ hardware behind their [USP or VSP] controller, and [IBM’s] SVC … allows you to take a number of different devices and virtualise them.

The virtualised option gives you a pool of storage from which you can provision but doesn’t necessarily meet all of the cloud storage requirements of scalability, multi-tenancy and chargeback. That’s the difference between the two, and clearly virtualised storage could be used in a cloud environment. Is there any way to transition from virtualised storage to an internal cloud, or is rip-and-replace the only option?

Evans: We could take an existing virtualised environment and transition that into a private cloud storage infrastructure. [We] would be looking to make sure … that we added additional features such as management and reporting. We’d want to be able to look at that environment and … carve it up by customer, report on usage by customer and scale it. That’s perfectly possible with a virtualised environment as long as we add those features on.

Some customers might decide their environments aren’t suitable for that and might want to do a rip-and-replace and put in new technology. There are storage arrays today that are more suited to [private] cloud deployments. Companies such as 3PAR have designed new infrastructures that scale better, that do multi-tenancy better and so on, and these may be a good choice for some customers.

There are also other projects, such as [EMC’s Atmos, which was] originally sold into service providers. But, a customer could deploy that in their own environment as a basis for private cloud storage following the same rules of scalability, multi-tenancy and so on.

Read more on Cloud storage